The effects of inclusion of different levels of raw kidney bean (Phaseolus vulgaris) of high lectin content (27 g/kg meal) in a high-quality (lactalbumin) control diet were tested in nutritional trials on the growth and metabolism of obese Zucker (fafa) rats and their lean littermates in comparison with pair-fed controls. All diets contained 100 g total protein/kg and either 50 g lipids/kg (low fat) or 150 g lipids/kg (moderate fat). The growth of both obese and lean rats on bean diets was retarded by the daily bean intake in a dose-dependent manner. However, most of this was because bean-fed rats contained less body fat than the controls after 10 d. Thus, after feeding low-fat diets containing up to 130 g kidney bean/kg (lectin intake ≤ 0·2 g/kg body weight (BW) per d) in both 10 d and 70 d trials, the bodies of obese rats contained less fat but not protein than their pair-fed controls. Moreover, by increasing the lipid content of the diet to 150 g/kg, the level of bean inclusion could be increased to 280 g/kg (lectin intake ≥ 0·4 g/kg BW per d) without loss of body protein and skeletal muscle. Although these rats contained more body fat than those which were fed on low-fat diets, their weight reduction could be accounted for exclusively by reduced lipid content. In contrast, significant body protein loss occurred when the same diet of high lectin content was fed to lean littermates. Plasma insulin levels were significantly depressed in the obese Zucker rats on bean diets but the pancreas was not significantly enlarged nor its insulin content changed in 10 d trials. However, significant pancreatic growth occurred on long-term (70 d) bean feeding compared with pair-fed controls. The results suggest that, in addition to animal nutrition, it may also be possible to use the bean lectin as a dietary adjunct or therapeutic agent to stimulate gut function and ameliorate obesity if a safe and effective dose-range can be established for human subjects.